1. Field of the Invention
The present invention pertains to methods for determining sources of interference, particularly sources of interference in a magnetic resonance apparatus that cause partial discharges in an encapsulated conductor structure.
2. Discussion of the Related Art
In magnetic resonance equipment generally, dynamic magnetic fields having linear gradients are superimposed on the static magnetic field in all three spatial directions. A gradient coil provided with at least three separate conductors, i.e., coils, is used. In order to achieve a desired compact construction, a high degree of mechanical strength and a high voltage strength, the intermediate spaces of the gradient coil are encapsulated with an insulating material, such as a flexible bisphenol-A system having anhydride hardeners.
A gradient coil typically has a diameter of approximately 100 cm, while the length dimensions can fluctuate between 120 and 200 cm. Based upon the encapsulation volumes resulting therefrom, with a typical reaction resin compound requirement of between 300 and 600 kg, it is obvious that during casting flaws, such as bubbles, shrink between 300 and 600 kg, it is obvious that during casting flaws, such as bubbles, shrink holes and cracks, will often arise that significantly decrease the voltage strength. These flaws form locations for the occurrence of partial discharges, whose probability increases as the voltage increases.
A basic method and apparatus for partial discharge measurement is described in DE 198 02 551, which shows that the apparatus is arranged on the lower coil, from which the partial discharge can be measured.
Partial discharges are partial breakdowns in a dielectric, triggered by a local field strength peak, which can for example be caused at the cited flaws in the insulating material. Since during operation, the gradient coils are charged with a high voltage in the range of several kV, a gradient coil having flaws is expected to have such partial discharges. In general, the threshold voltage of this partial discharge is thereby lower than the breakdown or dielectric strength of the dielectric material.
The actual risk to a component presented by partial discharges lies in the occurrence of irreversible destruction in sub-regions of the dielectric, and in the increased probability of a complete breakdown. However, such partial discharges also have a disadvantageous effect on the image exposure process, because image disturbances can for example be caused in the receive system of a magnetic resonance apparatus. In the recorded resonance image, these image disturbances are expressed in the form of what are known as spikes.
Conductor structures of this sort, in which it is highly probable that sources of interference will arise during manufacture, are first subjected to a quality and function test before their final installation in an apparatus, for example a magnetic resonance apparatus. The purpose of the test is to discover whether the conductor structures will operate in problem-free fashion or whether the conductor structures will tend to generate image disturbances. During the test, a high voltage is hereby applied to a conductor of the conductor structure, while a second, adjacent conductor, separated via an insulating layer in which interference points may be present, is connected to ground via a four-terminal coupling circuit.
Due to the adjacent high voltage, it is possible, using a partial discharge measurement apparatus, to measure signals in the relevant insulating layer that are coupled out from the high-voltage circuit and supplied to the partial discharge measuring apparatus. These signals are subsequently evaluated in order to discover whether they are partial discharge signals or whether the signals are merely located in the range of standard noise. The partial discharge measurement is thereby carried out in broadband fashion, in a frequency range from approximately 40 to 400 kHz. If points of interference are present in the encapsulation compound that tend towards partial discharges with a voltage that is below or in the range of the adjacent high voltage. Thus, interference points are actually excited to the point of partial discharge and can be detected, according to the method of the present invention. The adjacent high voltage is preferably located in the range of the standard operating voltage of the conductor structure.
Using the method of the present invention, a sufficient coarse qualification can be carried out in a conductor structure in order to discover whether or not it tends towards partial discharges. With respect to the specified dimensions of a gradient coil, it is clear that a relatively large number of coils comprise points of interference, and therefore exhibit partial discharges in the frequency range under examination so that the percentage of rejections is relatively large.
In order to solve this problem, the present invention involves measuring the specified first partial discharge, after which, dependent on the result of the first analysis, the conductor is again charged with a low-frequency high voltage, and the signals are measured within a frequency range located in the MHz range, using a partial discharge measurement apparatus. The signals are then analyzed in order to determine partial discharges.
According to the present invention, a second partial discharge measurement is executed in which the conductor is again placed adjacent to low-frequency high voltage. However, in this second measurement, the signals are measured and analyzed in the MHz range. Since the relevant image signals, recorded using a magnetic resonance apparatus, likewise exhibit frequencies in the MHz range, the second measurement step allows testing even though the partial discharges determined in the first measurement step also have high-frequency portions in the MHz range, i.e., in the imaging frequency range. Thus, image disturbances during operation would also be revealed.
This is because a partial discharge found during the first examination, and whose produced signal comprises a frequency portion in the range between 40 and 400 kHz, does not always also have a frequency portion in the MHz range, which would be the actual disturbing portion. A gradient coil is therefore qualified as unusable only if partial discharge signals are determined in the MHz range that would actually lead to image disturbances. Since this is not always the case despite the presence of partial discharge signals acquired in the first measurement pass, gradient coils that would still have been rejected as unusable after the first measurement can advantageously be qualified as usable.
The partial discharge signals can thereby be measured and analyzed in a frequency range from 10 to 300 MHz, and can be measured and analyzed with particular advantage in a range from 60 to 65 MHz. The frequency of the high voltage is between 10 Hz and several kHz, and is preferably 50 Hz.
According to the present invention, two high-voltage partial discharge measurements are carried out in different frequency ranges, and the second measurement takes place in a frequency range that is relevant for imaging. The adjacent effective high voltage is thereby in a range that corresponds to the standard operating voltage of a gradient coil. In both measurements, only a high voltage is applied that falls off in quasi-static fashion between the two conductors involved, which is the insulation path to be measured. The same voltage is therefore adjacent over the entire length of the conductor, and thus the gradient coil between the two conductors.
However, this state does not correspond to the state during operation of the conductor structure. This is because during operation, current is driven via the conductors. This flow of current has the result that the voltage falls off over the length of the conductor, hence a quasi-dynamic change in voltage results over the length of the conductor. However, if the voltage falls off over the path, the possibility arises that precisely at the location at which an interference point was found in the previous static measurement in the MHz range under current load, a voltage is adjacent that is too low for the partial discharge voltage to be reached. As a result, the partial discharge determined in the static case does not occur. Such a gradient coil, which would have been qualified as unusable after the second measurement pass, can be used nonetheless, since partial discharges will not occur in operation.
Dependent on the result of the second analysis, a source of current is connected to the conductor and a current flows via the conductor, such the partial discharge signals produced as a result of a voltage gradient built up via the coil are measured and analyzed using a partial discharge measurement apparatus. The partial discharge measurement apparatus is operated within a frequency range located in the MHz range, in particular in the frequency range from 10 to 100 MHz, preferably between 60 and 65 MHz, and a preferable current of several 100 Amperes flows through the conductor. The operating state of the gradient coil is very closely approximated, so that, after execution of this third testing step, a well-founded prediction can be made as to whether or not the conductor structure will breakdown during operation.
According to the present invention, the frequency of the adjacent high voltage during the measurement under current load should be located in the range between 10 Hz and several kHz, in particular at 1 kHz. In general, the adjacent high voltage should have a magnitude from a few hundred volts up to several kV, in particular 3 kV. This is to be understood as the effective voltage.
As a result of the execution of the second, and if necessary, the third measurement steps, the method of the present invention provides a significantly more sensitive quality test that is closer to operating conditions. Independent of whether only the second step or the third measurement step is carried out if, dependent on the second or on the third analysis result, one or more measurements can be carried out in order to determine the position in the encapsulation of an interference point that is causing a partial discharge. This is due to the fact that if the respective analysis result permits the conclusion that only a few interference points are present and removal of these appears to be possible, then the additional measurement for determining the position of the interference point supplies information as to the exact location of this interference point in the encapsulation.
If the position of the interference point is known, an attempt can be made to repair the gradient coil, which in any case is presumably unusable. For the case in which the position can be determined unambiguously and a repair is successfully carried out, the gradient coil can be used, otherwise it remains rejected. First, a rough localization of the flaw can take place using the measurement under current load, by changing the direction of flow, and by using differences in the partial discharge inception voltages to draw conclusions concerning the possible position of the flaw point. Subsequently, via a calculation of the voltage distribution or electrical field distribution, the points having high voltage load can be identified. In such a step, the localization method can then be used in locally limited fashion.
Such localization methods, which can be performed with the use of a plurality of suitable sensors arranged on the gradient coil, include acoustic localization of sound waves produced by a partial discharge source, localization of magnetic and electromagnetic fields produced by a partial discharge, or measurement of runtime differences of partial discharge signals along conductors at the conductor structure operated under high voltage.
Another useful localization method makes use of an x-ray beam directed onto the gradient coil. Using a punctiform x-ray beam, the entire surface of the gradient coil is to be scanned. At the same time, a voltage that is smaller than the voltage for partial discharges is applied to the insulating layer. No partial discharge signals can be measured. If the x-ray beam comes into contact with a partial discharge source, such as a flaw or hollow space. Then, due to the energy of the x-ray beam, ionization of the charge bearers in the interference point is promoted and a measurable partial discharge is caused. In this way, the flaw can be localized in geometrically precise fashion.
It is an object of the present invention to provide a method for determining sources of interference in a magnetic resonance apparatus that cause partial discharges.
It is another object of the present invention to provide a method that enables a more sensitive testing of partial discharges in an encapsulated conductor structure.